A neural network approach for predicting the structural behavior of concrete slabs

Tully, Susan Hentschel (1997) A neural network approach for predicting the structural behavior of concrete slabs. Masters thesis, Memorial University of Newfoundland.

[img] [English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.

Download (14MB)
  • [img] [English] PDF - Accepted Version
    Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.
    (Original Version)

Abstract

Reinforced concrete slabs exhibit complexities in their structural behavior due to the composite nature of the material and the multitude and variety of factors that affect such behavior. As such, current methods for the design and analysis of reinforced concrete slabs are limited in scope and are approximate at best as they must rely on the results of experimental tests, which are both costly and time-consuming to perform. The research embodied by this document investigates the use of a branch of artificial intelligence known as Neural Networks (NN) as a quick and reliable alternative to such experimental testing. -- Four neural network models are developed to predict the following aspects of the overall behavior of a concrete slab: 1) load-deflection behavior; 2) crack pattern at failure; 3) concrete strain distribution; and 4) reinforcing steel strain distribution. Results from experimental tests on thirty-four full scale slabs are utilized to develop these four models, incorporating all of the parameters that govern their behavior. The rationale behind and the details involved are explained for the setup, computer implementation and selection of each optimum neural network model. Results show that the neural network technique can perform as a satisfactory alternative to experimental testing or detailed calculations to provide speedy predictions of all four aspects of the structural behavior of concrete slabs. A comprehensive spreadsheet tool is next created to incorporate all four of the optimum neural networks. The spreadsheet uses readily available software and can be used by structural engineers for instantaneous access to the prediction of any or all of the four aspects of a concrete slab's behavior given minimal data to describe the slab and the loading conditions. This tool, combined with the results for the four neural network models, demonstrates the powerful capabilities and success of neural networks in the realm of civil and structural engineering in general and reinforced concrete design in particular. This approach could readily be expanded to include the same predictions for other structural concrete elements such as beams and shear walls.

Item Type: Thesis (Masters)
URI: http://research.library.mun.ca/id/eprint/5319
Item ID: 5319
Additional Information: Bibliography: leaves 103-108.
Department(s): Engineering and Applied Science, Faculty of
Date: 1997
Date Type: Submission
Library of Congress Subject Heading: Concrete slabs; Structural analysis (Engineering)--Computer programs; Neural networks (Computer science)

Actions (login required)

View Item View Item

Downloads

Downloads per month over the past year

View more statistics